In dopaminergic neurons, a-synuclein (aS) partitions between a disordered cytosolic state and a lipid-bound state. Binding of aS to membrane phospholipids is implicated in its functional role of synaptic regulation, but also impacts fibril formation associated with Parkinson's disease. Low concentrations of the histological dye Congo red (CR) previously have been shown to promote alpha-synuclein (aS) aggregation, whereas at higher levels the dye has an inhibitory effect on fibril formation. We have carried out a detailed NMR study of the interaction between aS and CR at pH 6, over a range of stoichiometries and concentrations, and show similarities as well as key differences relative to prior reports on binding of aS to lipid vesicles. NMR relaxation dispersion measurements yield the bound fraction and time scales for the interaction with CR. At a aS:CR molar ratio of 1:1, only a small fraction of aS (ca 3%) is bound to CR, the latter forming relatively large (> 100 kDa) micelles, but results in strong attenuation of the NMR signals of the N-terminal 60 residues of aS. Rapid exchange (kex 3000 s-1) between the free and the CR-bound state broadens resonances of aS by two processes: a magnetic field dependent contribution, caused by the chemical shift difference between free and bound states, and a nearly field-independent contribution caused by slower tumbling of aS bound to the CR micelle. Measurement of 13Ca secondary chemical shifts upon interaction with CR show changes for the interacting N-terminal residues (1-60) that are much smaller than expected if the protein were to adopt a pure alpha-helical geometry in the bound state, whereas amide-amide NOE intensities and the ratio of intra-residue and sequential HN-Ha NOEs point to less extended structure in the bound state than in free aS. At the same time, the absence of Ha(i)-HN(i+3) NOEs confirms the lack of significant population of alpha-helical structure. NOE and chemical shift data are compatible with a heterogeneous binding mode in which turn conformations are highly populated. Relaxation and chemical shift data indicate that the NAC region of aS (encompassing residues 60-90) is not involved in CR binding at CR:aS stoichiometries lower than ca. 1:1, mostly caused by competition for CR-binding by the higher affinity of the N-terminal 60 residues. At a 5:1 aS:CR molar ratio, the NAC region of aS also shows significant exchange broadening, with intramolecular sequential and intraresidue NOEs and 13Ca chemical shifts indicating again a heterogeneous binding mode that only shows a small increase in population of the alpha region of Ramachandran phi/psi space. A WATER-LOGSY experiment provides intermolecular NOE contact information between CR and aS, showing strongest contacts to a mostly hydrophobic subset of the N-terminal 30 residues and weaker, relatively uniform magnetization transfer to residues 40-100, with no interactions for the C-terminal 35 residues. A 2011 study by Selkoe et al reported that if aS is expressed in mammalian cells and purified without a heat denaturation step, it adopts a stable tetrameric helical structure. We developed this expression system but were unable to duplicate their findings. However, the N-terminal acetylation that occurs in mammalian cells does introduce transient helical structure (15% population) for the first 6 -terminal residues, which we found to lead to a pronounced increase in lipid binding affinity.